GLOBALISATION, EXPORT-ORIENTED EMPLOYMENT FOR WOMEN

Transcription

1 United Nations Research Institute for Social Development (UNRISD) GLOBALISATION, EXPORT-ORIENTED EMPLOYMENT FOR WOMEN AND SOCIAL POLICY: A CASE STUDY OF INDIA Paper prepared for the UNRISD project on Globalization, Export- Oriented Employment for Women and Social Policy ***DRAFT NOT FOR CITATION*** Jayati Ghosh Jawaharlal Nehru University New Delhi July

2 The United Nations Research Institute for Social Development (UNRISD) is an autonomous agency engaging in multidisciplinary research on the social dimensions of contemporary problems affecting development. Its work is guided by the conviction that, for effective development policies to be formulated, an understanding of the social and political context is crucial. The Institute attempts to provide governments, development agencies, grassroots organizations and scholars with a better understanding of how development policies and processes of economic, social and environmental change affect different social groups. Working through an extensive network of national research centres, UNRISD aims to promote original research and strengthen research capacity in developing countries. Current research programmes include: Civil Society and Social Movements; Democracy, Governance and Human Rights; Identities, Conflict and Cohesion; Social Policy and Development; and Technology, Business and Society. A list of the Institute s free and priced publications can be obtained by contacting the Reference Centre. UNRISD, Palais des Nations 1211 Geneva 10, Switzerland Tel: (41 22) Fax: (41 22) info@unrisd.org Web: Copyright United Nations Research Institute for Social Development. This is not a formal UNRISD publication. The responsibility for opinions expressed in signed studies rests solely with their author(s), and availability on the UNRISD Web site ( does not constitute an endorsement by UNRISD of the opinions expressed in them. 2

3 Content I. WOMEN WORKERS AND EXPORT PRODUCTION IN ASIA: RECENT ISSUES AND TRENDS... 4 II. WOMEN S EMPLOYMENT IN THE INDIAN MANUFACTURING SECTOR, INCLUDING IN EXPORT EMPLOYMENT III. EXPORT PROCESSING ZONES...25 IV. MEASURES FOR SOCIAL PROTECTION OF WOMEN WORKERS REFERENCES

4 This paper seeks to examine the Indian experience with respect to women s employment in export-oriented manufacturing industry in the era of globalisation. It also considers the role of social policy in providing work and survival security to women, by first evaluating the effects of state policy, and then considering other attempts to ensure minimum security to women workers. The first section sets out some of the issues with respect to the feminisation of labour in export-oriented employment, and situates the discussion in the context of the experience of the high-exporting East Asian economies in the 1990s. The evidence pointing to a fall in the share of women in export-oriented manufacturing employment even before the onset of the East Asian crisis is considered, and the possible reasons for it are discussed. With this background, the next section briefly highlights the important trends with respect to aggregate female employment in the Indian manufacturing sector over the 1990s. It is argued that much of the use of female labour in export production in India has been in informal and unorganised workplaces, including home-based work, with associated implications for pay, working conditions and consequently also for social policy. The cases of Export Oriented Units (EOUs) and Export Processing Zones (EPZs) are then taken up in the third section, with specific attention to what such employment has meant for job, material and social security. Issues relating to social protection of female labour through the agency of the state and other examples of attempts to provide social security are considered in the final section. In this section there is also an argument for the need to have a macroeconomic perspective on the conditions for improving employment conditions for women workers, which would have wider applicability to other developing countries as well. I. Women workers and export production in Asia: Recent issues and trends The link between export employment and the feminisation of employment is now well known. (See Horton [1995], Wee [1998] and Joekes [1999] for discussions of some of this literature). While feminisation of employment can refer to either the absolute or the relative increase in numbers of women employed, most of the literature on this process in export-oriented employment has tended to look at the share of women to total workers in particular sectors. This is because the absolute increase in such employment (or even an increase in the share of women so employed to total female labour force) could be part of a fairly standard development pattern whereby more and more people are drawn into labour markets determined by changing patterns of labour demand, but need not tell us anything about any particular preference for women workers. By contrast, the relative increase in the share of women in total export employment, which was so marked for a period in parts of Asia, is a qualitatively different phenomenon. Of course, such feminisation has obviously been reflected in more and more women being drawn into paid employment. This process was most marked over the period 1980 to 1995 in the high-exporting economies of East and Southeast Asia, where the share of female employment in total employment in the Export Processing Zones (EPZs) and export-oriented manufacturing industries typically exceeded 70 per cent. It was also observed in a number of other developing countries,for example in Latin America in certain types of export manufacture. Women workers were preferred by employers in export activities primarily because of the inferior conditions of work and pay that they were usually willing to accept. Thus, women workers had lower reservation wages than their male counterparts, were more willing to accept longer hours and unpleasant and often unhealthy or hazardous factory conditions, typically did not unionise or engage in other forms of collective bargaining to improve conditions, and did not ask for permanent contracts. They were thus easier to hire and fire at will and according to external demand conditions, and also, life cycle changes such as marriage and childbirth could be used as proximate causes to terminate employment. 4

5 Another important reason for feminisation was the greater flexibility afforded by such labour for employers, in terms of less secure contracts. Further, in certain of the newer sunrise industries of the period such as the computer hardware and consumer electronics sectors, the nature of the assembly line work - repetitive and detailed, with an emphasis on manual dexterity and fineness of elaboration - was felt to be especially suited to women. The high burnout associated with some of these activities meant that employers preferred work-forces that could be periodically replaced, which was easier when the employed group consisted of young women who could move on to other phases of their life cycle. The feminisation of such activities has had both positive and negative effects for the women concerned. On the one hand, it definitely meant greater recognition and remuneration of women s work, and typically improved the relative status and bargaining power of women within households, as well as their own self-worth, thereby leading to empowerment. [Such positive effects are documented in Heyzer [1988], Joekes and Weston [1995] and Kibria [1995], inter alia.] On the other hand, it is also true that most women are rarely if ever unemployed in their lives, in that they are almost continuously involved in various forms of productive or reproductive activities, even if they are not recognised as working or paid for such activities. This means that the increase in paid employment may lead to an onerous double burden of work unless other social policies and institutions emerge to deal with the work traditionally assigned to (unpaid) women. For example, without adequate socially provided crèche and child care facilities, or adequate and accessible medical care and hospitalisation, the job of looking after the young, the sick and the old, which is typically unpaid labour performed by women in a household, can devolve on girl children if the adult women are employed outside the home. Similarly the burden of regular housework typically continues even for women employed outside the home, except to the extent that these can be devolved to other household members or shifted to paid services. Given these features, it has been fairly clear for some time now that the feminisation of work need not be a cause for unqualified celebration on the part of those interested in improving women s material status. On the other hand, it is also well known that the very process of feminisation can also set in place social and political changes which improve the bargaining position of women not only within their own households, but also within the society and economy in general. The exposure to paid employment has also played a major role in encouraging greater social recognition of women s unpaid work and led to greater social pressure for improving the conditions of all work performed by women in a number of countries. As more and more women get drawn into the paid work force, there is greater public and social pressure generally for improvement in their conditions of work and security of contract, for greater health and safety regulation in the workplace, and for improvement in relative wages. Thus there are several reasons why, despite the acknowledged inferior conditions of such work, such a process of feminisation in labour markets was generally welcomed by the women who were involved in it. However, it is now becoming evident that the process of feminisation of labour in exportoriented industries may have been even more dependent upon the relative inferiority of remuneration and working conditions, than was generally supposed. This becomes very clear from a consideration of the pattern of female involvement in paid labour markets in East and Southeast Asia, and more specifically in the export industries, over the entire 1990s. It is well known that the expansion in export production which fueled the economic boom in the East Asian region in the decade 1985 to 1995 was largely based on the growing use of women as wage workers. Indeed, the Asian export boom was driven by the productive contributions of Asian women in many different ways: in the form of paid labour in export-related activities and in services, through the remittances made by migrant women workers, and through the vast amounts of unpaid labour of women as liberalisation and government fiscal contraction transferred many areas of public provision of goods and services to households (and thereby to women within households). 5

6 Most countries of the region (barring a few important exceptions notably India) the period between 1985 and 1997 witnessed a massive increase in the labour force participation of women. This process was most marked in the Southeast Asian region which was also the most dynamic in terms of exporting. Throughout Asia, as a consequence, the gap between male and female labour force participation rates narrowed, suggesting that this period was one in which - at least in terms of quantitative involvement - the gender gap narrowed. Indeed, this narrowing of the gender gap was not confined to overall employment it also extended into wage differentials and even working conditions as the proportion of women involved in such activities grew. In the exporting economies of Southeast Asia, these pressures were quite apparent, even if not always effective, from the early 1990s onwards, and to some limited extent they did contribute to a slight narrowing of the wage gap. But it is now evident that, as this more positive process occurred, there was in fact a decline in the share of women employed in the export manufacturing sectors. Thus, as the relative effective remuneration of women improved (in terms of the total package of wage and work and contract conditions), their attractiveness to employers decreased. This is discussed in a little more detail below, and is based on a more extended discussion of the issue in Ghosh [1999b]. Most observers would not be surprised to find that the share of women in employment in the East Asian region has fallen in the very recent past, since this is after all a pattern well noted in economic downturns. It is obvious that the crash of mid-1997 dramatically altered both the potential for continued economic activity at the pre-crisis rate, as well the conditions of employment in the East and Southeast Asian region. When the export industries started to slow down from the middle of 1995, it became evident that continued growth of employment in these export-oriented industries could not be the same engine of expansion that they had served as over the previous decade. Obviously, therefore, there could be some setback to the feminisation of employment that had been occurring, since the export industries had become the most important employers of women at the margin, especially in the large employment sectors such as textiles. Indeed, the very features which had made women workers more attractive to employers - the flexibility of hiring and firing and the more casual, non-unionised nature of labour contracts - are precisely those which are likely to render them to be the first to lose their jobs in any recessionary phase. But in fact the reduced role for women workers (at the margin) was something that was coming into play even before the effects of the economic crisis worked themselves through. It is now apparent that even the earlier common assessment of the feminisation of work in East Asia had been based on what was perhaps an overoptimistic expectation of expansion in female employment. Trends in aggregate manufacturing employment and female employment in the export manufacutring sector over the 1990s in some of the more important Southeast Asian countries, as described in Table 1, reveal at least two points of some significance. The first is that there is no clear picture of continuous employment in manufacturing industry over the decade even before the period of crisis. In several of these economies - South Korea, Singapore, Hong Kong China - aggregate manufacturing employment over the 1990s actually declined. Only in Malaysia, Indonesia and Thailand was there a definite upward trend to such employment. 6

7 Table 1: Trends in manufacturing employment and share of women workers Country and Year Total manufacturing employment, 000s Women employed in manufacturing, 000s Share of women workers, per cent South Korea, ,828 1, South Korea, ,474 1, Malaysia, , Malaysia, , Indonesia, ,693 3, Indonesia, ,773 4, Thailand, ,133 1, Thailand, ,334 2, Singapore, Singapore, Hong Kong SAR, Hong Kong SAR, Source: ILO Yearbook of Labour Statistics, 1994 & 1998 Some may see the trend of reduction of manufacturing employment as a typical indication of a mature economy, that is one in which the service sectors are achieving greater dominance and therefore there is a shift of labour away from secondary activities and towards a range of services. But remember that these are economies whose economic dynamism was fundamentally based on the ability to push out ever increasing quantities of manufacturing exports. And this reliance on exports was such that it fed into expectations in the rest of the economy, most particularly in the financial sector, thus creating conditions which made the crash possible. More significantly, it is precisely the ability of the manufacturing sector to respond (either to renewed export demand or to increased domestic demand resulting from a positive fiscal stimulus) which has determined the ability of the Southeast Asian economies to recover from crisis (Ghosh and Chandrasekhar [2001]). Thus, South Korea and Malaysia experienced partial recoveries which allowed both the volume of economic activity and employment to rebound led by the recovery in manufacturing activity. However Thailand has still not recovered even to pre-crisis activity levels, essentially because manufacturing growth has not picked up sufficiently. The renewed fears of stagnation, recession and possibly another financial shock that became widespread in the region in mid-2001 were again based on the slowdown in manufacturing activity. All this clearly points to economies which are still very much dependent upon increased manufacturing output as the basic reflection of economic expansion, which is quite far from the mature economy situation. The second point to emerge from the data over the 1990s is that, while they do show that female employment in manufacturing was important, the trend over the 1990s, even before the crash, was not necessarily upward. In most of the countries mentioned, there is a definite tendency towards a decline in the share of women workers in total manufacturing employment over the latter part of the 1990s. In Hong Kong and South Korea, the decline in female employment in manufacturing was even sharper than that in aggregate employment. Similarly, even in the countries in which aggregate manufacturing employment increases over the period , the female share has a tendency to stabilise or even fall. Thus, in Indonesia the share of women workers in all manufacturing sector workers increases from an admittedly high 45 per cent to as much as 47 per cent by 1993, and then falls to 44 per cent by In Malaysia the decline in female share is even sharper than in South Korea: from 47 per cent in 1992 to only 40 per cent in A slight decline is evident even in Thailand. This fall in women s share of employment is evident not just for total manufacturing but even for export-oriented manufacturing, and is corroborated by evidence from other sources. Thus Joekes [1999] shows that the share of women employed in EPZs declined even between 1980 and 1990 in 7

8 Malaysia, South Korea and the Philippines, with the decline being as sharp as more than 20 percentage points (from 75 per cent to only 54 per cent) in the case of Malaysia. In other words, what the evidence suggests is that the process of feminisation of export employment really peaked somewhere in the early 1990s (if not earlier in some countries) and that thereafter the process was not only less marked, but may even have begun to peter out. This is significant because it refers very clearly to the period before the effects of the financial crisis began to make themselves felt on real economic activity, and even before the slowdown in the growth rate of export production. So, while the crisis may have hastened the process whereby women workers are disproportionately prone to job loss because of the very nature of their employment contracts, in fact the marginal reliance on women workers in export manufacturing activity (or rather in the manufacturing sector in general) had already begun to reduce before the crisis. This is an important issue that clearly requires further investigation. The reversal of the process of feminisation of work has already been observed in other parts of the developing world, notably in Latin America. [ILO, 1998] Thus Ghiara [1999] points out that in Mexico, as the share of exports in the machinery sector increased between 1987 and 1993, the proportion of women employed fell from 38 per cent to 29 per cent. Quite often, such declines in female share of employment have been found to be associated with either one of two conditions: an overall decline in employment opportunities because of recession or structural adjustment measures, or a shift in the nature of the new employment generation towards more skilled or lucrative activities. In the East Asian case, until 1996 at least the first factor would not have been important, and while the second factor is certainly likely to have played a role, it would not have explained the entire shift that can be witnessed. Also, the shift towards more skilled activities was more marked in certain countires such as South Korea, Singapore and Malaysia, and less evident in others such as Thailand and the Philippines. But there may be another process which is associated with widespread feminisation of work, which creates conditions for its own unravelling over time. This relates to the relative cost of hiring women workers, and the relation to the perceived other advantages. As mentioned above, one of the important reasons for preferring women workers in many export-related activities in particular, has been the lower reservation and offer wages of women. Throughout the East Asian region, women workers wages have been consistently and significantly lower than male wages in the aggregate. The differentials have been particularly sharp in the case of South Korea, Malaysia and Singapore, where the average female wages were typically just above half those paid to male workers, as Table 2 indicates. Table 2: Female wages as per cent of male wages in manufacturing work Average Per Cent Bangladesh 71.7 Hongkong 65.9 South Korea 52.3 Malaysia 57.9 Philippines 84.0 Singapore 57.1 Sri Lanka 87.8 Thailand 63.8 Note: The data refer to an average of years for which data were available in the 1990s. Source: ILO Yearbook of Labour Statistics, 1998 To some extent this reflects differences in the types of jobs for which women are used, which are typically at the lower skill and lower wage end of the employment spectrum. But it also reflects the 8

9 general tendency for gender discrimination in pay, whereby women are paid less even for similar or identical jobs. This feature, which was quite marked in East Asia, is precisely the feature which tends to be rendered less potent as more and more women are drawn into employment. As women become an established part of the paid work force, and even the dominant part in certain sectors (as indeed they have become in the textiles, ready made garments and consumer electronics sectors of East Asia) it becomes more difficult to exercise the traditional type of gender discrimination at work. Not only is there an upward pressure on their wages, but there are other pressures for legislation which would improve their overall conditions of work. It is worth noting that the female wage as a proportion of the male wage has been rising in most parts of East Asia in the 1990s up to Thus, in South Korea, the ratio of average female wages to male wages increased from 50 per cent in 1990 to 56 per cent in 1997, while in Malaysia it moved from 49 per cent to 57 per cent between 1990 and In Thailand it improved from 63 per cent in 1991 to 68 per cent in In Singapore it went up from 54 per cent in 1990 to 60 per cent in Hong Kong is the only economy in the region for which there is evidence of a decline in this ratio over this period, from 69 per cent in 1990 to 61 per cent in [ILO 1999] What this narrowing of the wages gap has meant is that women became less cheap as labour in exporting industry. To add to this, over this period there have been several moves towards protecting the interests of women workers, for example in terms of slightly better maternity benefits and some improvement in the nature of contracts. In South Korea, a law which allowed women to be fired once they got married was repealed in the early part of this decade. Other legislation in other countries in the region has allowed for a modicum of benefits which were previously denied to women workers, to be provided. Thus, it has been observed that in several of these countries there were moves to ensure longer maternity leave of upto three months, provide better housing conditions and health care to young women workers and also (after the bad publicity offered by a series of industrial accidents) to work towards better compliance with minimal safety norms. [Lim 1994, 1996] While this is fundamentally necessary and desirable, such social action reduces the relative attractiveness of women workers for those employers who had earlier been relying on the inferior conditions of women s work to enhance their export profitability. The rise in wages also tends to have the same effect. If this is in fact one of the explanations for the tendency towards reduced employment of women in export activities in the region, then it raises certain crucial questions which will become increasingly important. How is it possible to ensure a minimum provision of basic rights and privileges to women workers, and to improve the conditions of their work, without simultaneously eroding their attractiveness to employers and reducing the extent of female wage employment? How can such rights and basic labour standards be assured in the coming phase, in which heightened export competition is likely to be combined with a phase of aggregate employment contraction, as the full force of the current adjustment measures and the slowdown of the world economy are felt in the region? 9

10 Table 3: Unemployment Rates Year Total % Male % Female % Indonesia Hong Kong Pakistan Philippines Philippines Singapore Sri Lanka Thailand Source: ILO Yearbook of Labour Statistics 1998 These are already pressing concerns, as is evident from the growth of unemployment in the region, which is bound to accelerate in the coming period. Already, over the 1990s, female open unemployment rates in the East Asian region were much higher than those of men even in the expansionary phase. This is clear from Table 3 which describes open unemployment rates before the onset of the crisis of But now, as the exporting industries are hit and as the general economic decline is worsened by adjustment measures which are moving towards a downward harmonisation of labour standards for all workers (men as well as women) the problem is likely to become more acute. Thus, in South Korea as a consequence of economic crisis, the fall in regular employment was much more severe for women than for men. Between 1997 and 1998, regular employment for men workers fell by 7 per cent, but that for women workers fell by 20 per cent. [Korean Working Women s Network, 1998] Out of the 47 per cent of Korean workers employed in temporary capacity in 1998, 55 per cent were women. In addition, the proportion of women who were defined as economically inactive increased by 6.4 per cent, indicating a strong discouraged worker effect. [Lee, 2001] Also, as male workers were effectively forced to accept worse employment conditions, the working conditions for all workers deteriorated even as less women found employment at the margin. This process naturally gained momentum as the rate of growth of exports decelerated across the region from mid-1995 onwards, but the important point is that the process had started well before that, in fact during the height of the economic boom in these countries. Indeed, with the onset of crisis and the recession of , there appears to have been a shift to more insecure very small unit-based or home-based employment of women workers, in production chains based on a substantial dependence on outsourcing by large final distributors. Already this was a prevalent tendency in the region. For example, labour flexibility surveys in the Philippines have shown that the greater the degree of labour casualisation, the higher the proportion of total employment consisting of women and the more vulnerable these women are to exploitative conditions. [ILO 1995] This became even more marked in the post-crisis adjustment phase. [Pabico 1999] In Southeast Asia, women have made up a significant proportion of the informal manufacturing industry workforce, in garment workshops, shoe factories and craft industries. Many women also carry out informal activities as temporary workers in farming or in the building industry. In Malaysia, over a third of all electronics, textile and garments firms were found to use sub-contracting. In Thailand, it has been estimated that as many as 38 per cent of clothing workers are homeworkers and the figure is said to be percent in the Philippines [Sethuraman 2000]. Home-based workers, working for their own account or on a subcontracting basis, have been found to make products ranging from clothing and footwear to artificial flowers, carpets, electronics and teleservices. [Carr and Chen, 1999; Lund and Srinivas 2000] 10

11 This is of course part of a wider international tendency of somewhat longer duration: the emergence of international suppliers of goods who rely less and less on direct production within a specific location and more on subcontracting a greater part of their production activities. Thus, the recent period has seen the emergence and market domination of manufacturers without factories, as multinational firms such as Nike and Adidas effectively rely on a complex system of outsourced and subcontracted production based on centrally determined design and quality control. It is true that the increasing use of outsourcing is not confined to export firms; however, because of the flexibility offered by subcontracting, it is clearly of even greater advantage in the intensely competitive exporting sectors and therefore tends to be even more widely used there. Much of this outsourcing activity is based in Asia, although Latin America is also emerging as an important location once again. [Bonacich et al., 1994] Such subcontracted producers in turn vary in size and manufacturing capacity, from medium-sized factories to pure middlemen collecting the output of home-based workers. The crucial role of women workers in such international production activity is now increasingly recognised, whether as wage labour in small factories and workshops run by subcontracting firms, or as piece-rate payment based homeworkers who deal with middlemen in a complex production chain. [Beneria and Roldan, 1987; Mejia, 1997] It has been suggested that a larger proportion of such subcontracting in fact extends down to homebased work. Thus, in the garments industry alone, the percentage of homeworkers to total workers has been estimated at 38 per cent in Thailand, between per cent in the Philippines, 30 per cent in one region of Mexico, between per cent in Chile and 45 per cent in Venezuela. [Chen, Sebstad and O'Connell, 1998] All this may have special significance in India, where the process of open feminisation of work is evident only in a very limited way in certain specific sectors and regions of export manufacturing. Despite the hype that is heard about the dynamic role of export employment, including in improving the conditions of women, there are few indications thus far that there has been any major shift in either the structure of production or female employment patterns. But there are important changes at the margin, which suggest that there is indeed a growing significance of female participation in exportoriented production in India, even if in ways rather different from the experience of the East Asian economies. II. Women s employment in the Indian manufacturing sector, including in export employment It is useful to begin by considering the overall evidence of the trends in female employment in urban India, and then focus on export-oriented activities, especially in manufacturing. A discussion on the recent developments in women s employment in IT-enabled services, an area with strong export potential, is also in order, and will be touched on briefly at the end of this section. In terms of aggregate urban employment, we have data from the small samples of the National Sample Survey ( a major survey carried out periodically by the Indian government s Central Statistical Organisation, dealing with employment and unemployment as well as with consumer expenditure in seprate surveys hereafter NSS) covering the period up to While these are not adequate to give us state-wise indicators, they do provide some idea of the tendencies at the all India level. All the results from most recent large survey for have still not been released and so will not be analysed here. It should be noted that the definition of economic activity used by the NSS is quite restrictive, and does not include the full spectrum of economic activities defined in the UN System of National Accounts, even though it now tries to take account of involvement in some household enterprises such as farm activities or small-scale artisan production or transacted service provision. It therefore excludes a significant amount of unpaid or non-marketed labour within the household, especially by women, including the processing of primary produce for own consumption, basic domestic handicraft 11

12 production, services such as cleaning, child care and so on, which are undertaken within the household and not marketed. This means there is a likely underestimation of economic activity within the household, as well as of the work participation rates especially of women. Given this caveat, Chart 1 provides an estimate of the overall work participation rate of urban women over the 1990s. Notwithstanding the amount of fluctuation, the overall trend appears to be one of decline. Indeed, the decline, on a point-to-point basis, is a very significant one for just a decade, by more than 4 percentage points from 15.2 per cent in to 11.4 per cent in This is extremely interesting, for it suggests that the picture that was being painted in the early 1990s, of a process of "feminisation" of employment, especially export-oriented manufacturing at the margin, has not been substantial enough to counteract other forces which have made for downward pressure on work participation rates. Per cent Chart 1: Work Participation Rate for Urban Women (Principal & subsidiary occupations) h h h1 Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. It also emerges that this decline in work participation has been associated with a rise in unemployment rates. It should be remembered that in countries like India, open unemployment rates are very poor indicators of the actual levels of job availability, because the material circumstances and absence of public social security systems mean that most workers have little choice but to find some employment, however unremunerative. Underemployment or disguised unemployment, which is far more difficult to estimate and measure, is therefore the most common tendency. It is true that there is more likely to be recognition of open unemployment in urban areas where surplus labour is less easily disguised as work in the fields. However, even here, the proliferation of informal activities, mostly in the service sector, can serve as a way of camouflaging the actual extent of underemployment. This is why the significant increase in open unemployment rates over the course of the 1990s, as evident from Chart 2, is quite remarkable. Already at the start of the decade, the rates were relatively high, indeed 50 per cent more than the average rates of the previous decade. Subsequently, they have dipped slightly and then come back to the previous high level of more than 8 per cent open unemployment. It is possible that this has actually been combined with the well-known discouraged worker effect, since aggregate labour force participation rates have also come down over this period. 12

13 Chart 2: Unemployment among Urban Women h1 % of workers % of popuation Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. Chart 3 shows that there is no evidence of shift towards employment in the secondary (or industrial) sector over the period since Indeed, there has been a slight decline in such employment relative to the total, while tertiary (services) sector employment has grown at the expense of both primary and secondary employment. Chart 3: Female Employment by sector h h h1 Primary Secondary Tertiary Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. A word about the nature of definitions of work activity is again in order. The NSS data on employment is based on the distinction between "principal" and "subsidiary" status of activity as well whether the person is "usually" engaged in the activity. Thus, a person is classified as "usual principal status" according to the status of the activity (or non-activity) on which the person spent a relatively longer time of the preceding year. The activities pursued by a person are grouped into three broad categories: (a) working or employed (b) seeking or available for work (i.e. unemployed) and (c) not in the labour force. 13

14 A "non-worker" (on the basis of the usual principal status) is someone whose major part of time in the preceding year was spent as either unemployed or not in the labour force. However, he or she could still be involved in some economic activity in a subsidiary capacity - when this is usually the case the person is referred to as a "subsidiary status worker". The two categories together - usual workers by both principal and subsidiary status - constitute "all usual workers". This detail can make quite a difference in the case of women workers - not only because much of their activity goes unrecognised, but because it is possible that they are classified as usually working when in fact it may reflect underemployment or engagement in a subsidiary activity only. Indeed we shall find that there can be substantial variation in the type of employment contract depending upon whether the activity is a principal one or a subsidiary one. Chart 4 is provided to allow for examining whether this makes any difference to the relative significance of manufacturing employment for urban women. It is possible, for example, that because of certain types of labour arrangements, women are utilised in manufacturing employment in more of a subsidiary status capacity. However, Chart 4 suggests that this is not actually the case, since the two indicators (principal status only and principal and subsidiary status taken together) move approximately together. Cha rt 4: Com pa rison of M a nufa cturing Em ploym ent by principal status only and PS+SS PS only PS+SS h h Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. The data on type of contract suggests that there has been a gradual shift away from selfemployment towards regular employment, with a share of casual employment fluctuation within the range of around per cent. This is evident from Chart 5. In itself, this a is positive sign, since usually in urban India, except for a few cases at the very high income levels, regular employment suggests higher wages and better conditions than either self-employment or casual work. 14

15 Chart 5: Distribution of women workers by type of employment Self-emp Regular Casual h h Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. Note, however, that this chart covers all workers, that is those with both principal and subsidiary status, and a more disaggregated look suggests that this may be misleading. Thus, as is clear from Chart 6, all workers together show a generally higher rate of self-employment than principal status workers only. And when only the latter are considered, then the decline in self-employment is much less substantial. Thus the shift away from self-employment could largely be a phenomenon that is significant only in terms of subsidiary activities, for a reason that will be considered later. Chart 6: Comparison of Self Employed by PS and PS+SS PS+SS PS only h h PS= Principal Status SS= Subsidiary Status Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. Similarly, consider the evidence of Chart 7: once again, regular employment by principal status alone is much less significant than it is when principal and subsidiary work is taken together, and again the increase is much less marked for principal activity alone than it is for all work. This suggests that much of the regular employment is occurring in terms of subsidiary status activity. By 15

16 contrast, as shown in Chart 8, casual employment remains much the same whether principal status is taken alone or whether all work is taken together. Chart 7: Comparison of regular employment by PS and PS+SS PS+SS PS only h h PS= Principal Status SS= Subsidiary Status Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. Chart 8: Comparison of casual employment by PS and PS+SS PS+SS PS only h h h1 PS= Principal Status SS= Subsidiary Status Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. So the big shift, apparently, in recognised work by urban women in India, is an increase in regular work in the form of subsidiary activity. What could be the form of this increase? It is argued in this paper that this has really taken two forms: the first is the increase in certain types of regular service activity, including domestic service; the second - and the important one for our purposes -is the increase in putting out home-based or other work as part of a subcontracting system for export and domestic manufacturing. Such work does not get incorporated in the employment statistics which are based on employers records, and this may explain the paradox that even while women s share of employment in manufacturing has not increased, the dependence of the sector - and especially of export-oriented manufacturing - on the productive contribution of women may well have increased. 16

17 This is considered below. But note first of all, that the composition of work and unemployment differs quite sharply across women from different household categories, as Chart 9 shows. The households here are categorised according to the work status of the head of household rather than the women workers in them. Predictably, female work participation rates are highest among casual labour households, who are typically the most likely to require outside earnings from as many family members as possible for sheer survival. But it turns out that unemployment rates are highest among women of regularly employed households, which is interesting. Chart 9: Work participation and unemployment rates by household type (1997) work partic unempl selfemployed regular empl casual labour others all Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. The data on women s employment in manufacturing in terms of category of enterprise confirms the picture that direct and formally recognised involvement of women has if anything come down in the period of the relative higher growth of exports in the early 1990s. This is shown in Chart 10, which reveals that there has been a fall in absolute numbers of women workers, as well as in Chart 11, which indicates a substantial decline in women's share of employment in Own Account Manufacturing Enterprises (OAMEs) in the period between and Employment in Non-Directory Manufacturing Establishments (NDMEs) (less than six workers) and in Directory Manufacturing Establishments (DMEs) (six or more workers) has remained broadly stagnant, which is itself significant considering that manufacturing output more than doubled over this period. Indeed, the overall stagnation of total employment in this area means that the absolute number of women employed in these enterprises declined. Chart 10: Female workers in unorganised manufacturing (numbers,000) OAMEs NDMEs DMEs Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. 17

18 Chart 11: Female workers in unorganised manufacturing (percentage) OAMEs NDMEs DMEs Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India. However, while overall women s employment in manufacturing has decreased over this period, there are important variations across particular industries. These can be gleaned from Table 4, which shows that the share of women employed and the trend over this period have varied quite remarkably across sub-sectors. In jute and beverages industries, the proportion of women was very high (as much as 80 per cent in jute) at the end of the 1980s, while in some others such as cotton and chemicals it was also quite high at nearly half the work force. Table 4: Female workers to total workers by sector of urban unorganised manufacturing (in percent) Food Beverages Cotton Jute Textiles Wood Paper Leather Chemicals Rubber Non-metals Basic metals Metal products Non-electricals Electricals Transport Other manuf All industries Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India, as quoted in Lalitha [1999] 18

19 Table 5: Share of women in total employment by type of manufacturing enterprise (in percent) Food OAMEs NDMEs DMEs Total Beverages Cotton Jute Textiles Chemicals Non-metals All industries Source: National Sample Survey Rounds on "Employment and Unemployment in India", Department of Statistics, Government of India, as quoted in Lalitha [1999] But the basic point remains that in most of these sectors the share of women declined over the first half of the 1990s, in some cases quite substantially. Only in the chemicals and non-metal manufacturing sectors was there any increase in the proportion of women employed. It is worth noting that these two sectors are also among the more hazardous industries. The involvement of women in own account small scale activity in the chemicals industry in particular may become problematic because of the numerous health hazards associated with such production especially (but obviously not exclusively) on a small scale. Table 5 shows the same data disaggregated across types of enterprise, for the sectors in which women s employment has been 20 per cent or more. It is evident that in most such industries, the decline of women in the share of employment has been most marked for Own Account Manufacturing Enterprises (which fits in with the pattern noted above) and Directory Manufacturing Establishments. It is being argued in this paper that home-based subcontracting activities, or work in very small units that do not even constitute manufactories, often on piece rate basis and usually very poorly paid and without any known non-wage benefits, may to some extent have substituted for both self-employment and more regular employment on a regular wage or salary basis. This is supported by some micro-evidence relating to certain manufacutirng sectors in particular towns and cities (Mukhopadhyay [1999]; Neetha [2001]; Shah and Gandhi [1998] Deshpande [2001]) found an 19

20 increase in putting out tailoring activities in three Mumbai slums. Das [2001] reports an increase in the involvement of women in unregistered piece-rate payment based manufacturing in the ceramicware industry in Gujarat. Anand [2001] has documented the pattern of integration of homebased work with wider marketing channels in handicrafts such as applique and coir products in parts of Orissa. Mahadevia [2001] finds an increase in the informalisation of employment and greater participation of women in homebased self-employment dealing with middlemen and contractors in Ahmedabad city of Gujarat, even though the city s role as a major centre of textile export is on the decline. Both the general pressure of industrial capitalist production and the particular external pressures faced by exporting industries which have to respond to international competition, operate to increase this tendency rather than to increase a more regular and secure form of women s involvement in manufacturing work. This perception is supported by the evidence on increase in subcontracting in Indian industry, especially in terms of export-oriented manufacturing companies and multinational companies operating in India. [Suri, ed., 1988; Bose, 1996; Ramaswamy, 1999.] One particular estimate of subcontracting intensity for Indian manufacturing (defined as the ratio of the value of goods sold in the same condition as purchased, to value added) found that it had increased from 9.46 per cent in 1970 to 25.3 per cent in , for all manufacturing sectors taken together. [Ramaswamy 1999, pp , also for following data.] Certain industries, especially consumer non-durable goods, were found to have very high subcontracting intensities in excess of 100 per cent by the early 1990s. These included weaving and finishing of cotton textiles on powerlooms (110 per cent), stationery articles (180 per cent) canning and preparation of fruits and vegetables (178 per cent) and even white goods such as refrigerators and air conditioners (115 per cent). Ratios in excess of 100 per cent indicate that the value of subcontracted production which forms part of the input, is even higher than the value addition involved in the final output. This particular study by Ramaswami [1999] identified both technology and labour regulation as factors behind the higher subcontracting activity of factories that used more labour per unit of output. Thus, technological changes that are oriented towards higher labour productivity obviously reduce the requirement of many kinds of unskilled labour in particular, but they also allow a break-up or physical disintegration of the production process which greatly facilities outsourcing. Similarly, labour regulation that puts requirements on worker pay, conditions or safety on employers who gather workers together in factory or workshop settings tends to encourage more use of putting out systems, since these are efffectively free from all such labour regulation and the chances of self-exploitation by homebased workers are greater. So, those activities or parts of production processes that involve more labour use per unit of output are more likely to result in organisational forms dependent upon subcontracting at various levels. This conclusion is also supported by other work on women workers in informal economic activities. Thus, for example, the well-known case of workers in the beedi (local leaf-based cigarette) industry, where approximately 90 per cent of the total workforce consisted of women and children working at home. [Labour Bureau, 1995] The beedi industry is not an export industry, but this suggests that home-based work is already a common manufacturing practice. These are not exportoriented units, but they indicate that such labour practices are widespread and therefore are available for use by export-oriented producers who naturally require even more flexibility in their functioning. Similarly, studies of the export-oriented industries of cashew and coir processing in Kerala found that male workers with higher reservation wages often stayed unemployed while women from the same household worked both at home and in small outside units at very low rates of remuneration. [Quoted in Mukhopadhyay, 1999.]. If the macro data combined with the micro evidence that is available are accurate representations of current trends, then the feminisation of export-oriented employment may have taken a particularly regressive form in India, whereby the marginal utilisation of women workers is at the lowest and poorest paid parts of the production chain, and such women are therefore effectively 20